OLa...

Cine1 shows more black compression than CINE4. Cine1 will bring out more highlight detail and slightly crush the blacks. Accordingly, I would use CINE1 on brighter scenes than CINE4. Cine4 will bring out more shadow detail than Cine1.

While the matrix color adjustments can be applied to any Cine curve preset, there will be considerable luminance variations between the cine gammas. If the mood of your film calls for these level changes in different scenes, by all means, use them. However, I don't think a brightness change would be consistent within a single scene, depending on how it's implemented. For example, using the transition buttons to go from A to B in a timed ramp, could be very effective.

OLa...

Cine1 shows more black compression than CINE4. Cine1 will bring out more highlight detail and slightly crush the blacks. Accordingly, I would use CINE1 on brighter scenes than CINE4. Cine4 will bring out more shadow detail than Cine1.

While the matrix color adjustments can be applied to any Cine curve preset, there will be considerable luminance variations between the cine gammas. If the mood of your film calls for these level changes in different scenes, by all means, use them. However, I don't think a brightness change would be consistent within a single scene, depending on how it's implemented. For example, using the transition buttons to go from A to B in a timed ramp, could be very effective.


Thanks for your quick reply Bill. I really enjoyed trying out your PP:s!!
Finally - for alround use - which gamma would be your choice? Best compromise? Cine3? And then switch to Cine1 on bright scenes and Cine4 when it is extremely murky?

/ola

Hi Bill,

I' m trying to understand better the theory behind colour matrix settings, which is not easy without a good (and calibrated) equipment. Please tell me wheter I understand your TC settings correctly:

Phase.................-5: why would you need to change that? May the default look different on a unit-to-unit basis?

R-G...................75
and
G-R................. -18: increase orange, substract yellow from green?

R-B.....................0
and
B-R...................-27: desaturate magenta from blue?

G-B...................-32
and
B-G.................. .13: increase blue and decrease greens in the cyan?

This is how I see it, please tell me whether I am right or not...

Also, another question: with all the pairs at default (zero) position, my unit's colours under regular home bulbs are definitely more accurate with the FL Light setting than anything else - why is that? Again, a unit-to-unit discrepancy? I'd think FL Light should work best under fluorescent lights...

PS: has anyone else noticed that - while of course much lower resolution - the EVF is actually much better for judging colour balance that the LCD? Perhaps it's just because you always look at it in the same way (no external light influence), or is it just my unit?

TIA

http://www.paolociccone.com/hd100-calibration.html

Thanks Bill, I saw that site. Paolo there uses slightly different terminology, so if you just commented with "Correct" or "Incorrect" my above interpretations of what the individual pairs are supposed to do, I'd appreciae it indeed.

When looking at the EX1 factory settings on the WFM, I found the colors to be inexact. My results lined up the color patches to the vectorscope targets much better than the factory settings. I rarely used "specialty" presets like FLLight and am not familiar with them. The EX1 matrix settings respond differently than the matrix settings on the HD110. My process was somewhat trial and error, not really understanding Sony's somewhat esoteric nomenclature. I assumed the 3 pairs were somehow related to Gain and rotation.

In my humble opinion, factory presets are consumer level conveniences that are not really useable. Nevertheless, there is more benefit to a casual user to go out and shoot, shoot, shoot than to spend hours dissecting the technical details. The EX1 is a very complex camera. Playing with settings without a WFM and vectorscope is inviting disaster. The human eye can't see critical details by looking at a viewfinder or monitor.

In any production process, especially a process that requires "tuning" or adjustment of the end product to meet certain QC requirements, the performance of any single unit is supposed to be within statistical variation of some median, usually 1 sigma. So, the answer to the question of whether there is unit to unit variation is really a function of how much money the manufacturer wanted to spend on QC. Generally speaking, a company like Sony needs to be fairly exact... with very small unit to unit variations.

Bill, the thread is becoming too long going back through every page. You used your own custom wedge, and colour chart, like DSC? Printing method? Thank you.

No....wrong!!
There is no way I could duplicate a DSC color chart. Read the material again. I used a DSC color chart.

I used my own B&W wedge. In this case, read the material again , as I made a disclaimer. The hi-lite/shadow levels were cross checked against the DCS chart.

Thanks, Bill. Sorry I never went back and deleted/corrected the question.

Bill just for an update,
what are you now using for your latest "True Color" profile?

Steve...

I still use a mix of TC2 with Black around -8. I prefer STD1 because it gives me the best latitude in post. If the scene is outdoors, in sunlight, STD1 blows out easily and I switch to Cine1 or cine4.

I'm experimenting with a CINEMA setting instead of HISAT. New color matrix. If you're interested, let me know.

I'm experimenting with a CINEMA setting instead of HISAT. New color matrix. If you're interested, let me know.

I guess we ALL are, Bill! Frankly (but this is limited to what I can see with my naked eye), the Cinema matrix offers much less saturated colours - even when compared to OFF, not to mention Hisat - which is a bit strange. I also find the FL Light matrix to be better balanced in terms of the hue, even under incandescent light - why is that?

Please keep us posted!

Piotr...

The EX1 seems to have the common digital camera sensor situation of being much more sensitive to reds than to blue hues. The FLLight preset boosts the red gain even further and I find it to be quite unacceptable. HISAT is a general, across the board boost in color saturation altho' the reds still dominate. My TC2 matrix is an attempt to get those reds to have less gain. The CINEMA mode boosts blue/cyan/green and attenuates the reds even further. So, with this as a starting point, I'm working new matrices along with a COLOR CORRECTION to boost cyan even further. Not for the feint of heart..;o)

If you like the FLLight, your eye must have a preference for warmer hues. I don't much care for that look.

You're right Bill - reds tend to dominate, but what I meant is that the reds in Hisat (unmodified) have too much yellowish/greenish tint to my liking (showing mostly in brown being too yellow-green).

I'm also trying to desaturate reds in favour of cyan, make browns less yellow, and get rid of magenta in blue. But without the proper equipment, so far I can't seem to achieve my goals...

I'm also trying to desaturate reds in favour of cyan, make browns less yellow, and get rid of magenta in blue. But without the proper equipment, so far I can't seem to achieve my goals...

I'm finding, with the vectorscope and DSC Labs color camette, that reducing the gain on the red/magenta and boosting the gain on the blue/cyan/green is very, very difficult. I've spent hours looking for the combination of matrix settings to get to this space; and, the best I've been able to do is TC2. My experiments with CINEMA have optimized the matrix by using COLOR CORRECTION mode in the cyan.Colors aren't as saturated, so I'm working with increased gain. It's a work in progress.

Thanks Bill for your continued work on the presets. I will be home this weekend after a two week shoot were I used your preset with cine 1 & 4 with success. It will be interesting to view on my larger monitors compared to the 17" Macbook pro.

So far all looks ok but I did change the BLacks to -4 an it seem to help in the high contrast shoots.

Always interested in what you find next.

Everyone...

What is the recipe for obtaining maximum tonal RANGE? Something that will work in every situation regardless if it's bright or in low light?

I would like to capture all my images with as much shadow AND highlight detail as possible, without crushed blacks or blown out whites. Something with a smooth and even tonal range from 0-108%.

I guess this would be called the "holy grail preset?"

warren...no such animal. no such thing as a perfect compromise.
latitude in hi contrast scenes is entirely different from latitude in low contrast scenes. anything that works well in low contrast will not work well in hi contrast and visa versa. throw the dice and takes your chance.

Everyone...

What is the recipe for obtaining maximum tonal RANGE? Something that will work in every situation regardless if it's bright or in low light?

I would like to capture all my images with as much shadow AND highlight detail as possible, without crushed blacks or blown out whites. Something with a smooth and even tonal range from 0-108%.

I guess this would be called the "holy grail preset?"

You need find a preset (PP) for each of the lighting conditions you shoot under. Generally you will loose a bit at one end but even so the latitude of this camera is amazing.

Here is something you might try that I do. I want the most detail in the blacks as I reasonably can, IE I don't want crushed blacks and CINE4 is a good preset for that but CINE4 is a very "bright" gamma and I shoot under the high hot sun, so I use CINE4 with the (black stretch slightly dropped) and under-expose by about 1.5 stops.

Mike

Thanks Bill & Mike!

@Bill:
I have a problem with your TC2-Preset. I get a greenish looking sky with your preset.

I did a comparison between your TC2 preset (with Black and Blackgamma=0), mine (hisat with Level=5 but no other modifications, Cine4, detail=off, Gamma=0, Black=0, Blackgamma=0) and a HV20 (TV-Mode, Aperture under my control - never thought the HV20 looks so bad compared to the EX1!)

http://rapidshare.com/files/93376520/EX1_TC2_Hisat_vs_HV20.mp4.html


regards Dennis

recheck your matrix values. you didn't set the to the right values I speced.

recheck your matrix values. you didn't set the to the right values I speced.


I used your hisat matrix values (TC2) with other gamma and black gamma settings because the blacks were crushed.

I cannot test with your original gamma and black gamma settings because it's rainy today. ;)


regards Dennis

check again

@Bill:
These were the settings I've used:

Matrix=High SAT
Phase=-5
R-G=+75
R-B=0
G-R=-18
G-B=-32
B-R=-27
B-G=+13

OffSet White=Off

Detail=Off

Skin Tone Detail=Off

Gamma Level=0
Select=Cine4

Black=0
Black Gamma=0

Low Key Sat=+10



regards Dennis

Did Bill update his settings? I don't remeber the +10 LowKey Sat and I have Black at -12. Results very nice.

Looks right, except:
Black: -12
Low Key Sat: 0

I have, since, backed off to Black at -8 to givre me a little more headroom in the shadows. I run these all settings with STD1, CINE1, CINE3, CINE4. Each one shifts the exposure latitude, depending on scene illumination.

STD1 gives a pretty contrasty image in the EVF, but, grades very well in post.

Be SURE you white balance.

The settings are perfect for sunny scenes. But what about night scenes outdoors in low light. Is there any way to get rid of the noise? I tried switching to STD 1 and this helped a bit getting rid of the noise but not completely. Is there something else I need to be doing?
Cody

Hmm, I've checked again, because wheather is great today.

It seems that the picture looks more natural with your preset.
(reddish sky with mine, greenish sky with Bill's TC2 but it looks better)


Edit: why do these pictures look different when viewing in firefox?
My preset seems to produce more natural colors when viewing in firefox - strange...

Edit2: The MXF files look the same in VLC compared to firefox, but very different in windows live photogallery...

regards Dennis

OK guys.

After some more testing, I must revise my understanding of what the philosophy behind both CINE gammas and CINE matrix is on this camera. Certainly, it's NOT the same I had on the V1, and thus NOT what I expected. Rather than more punchy/saturated than the STD gammas/ STD matrix, they seem to be muted and washet-out. BUT, given a second thought, do CINEmatic picture (as seen in the movie theaters) look contrasty? Not at all, so perhaps this is the CineAlta philosophy, after all? Low contrast, soft and saturated pictures...

Anyway, the two grabs below show a comparison of the same scenery using:

- Bill Raven's TC (or was it TC2), with Cine1 gamma on the left
- my new PP on the right:

gamma: STD1 level 0
matrix: Hisat level 20, phase 15, all colour pairs at zero (default)
Black: -25
Black gamma: -15
Low Key Sat: 15

As you can see, the STD1-based picture is much brighter and punchy than the CINE1-based one. What's more, this PP's knee (unlike that of CINE1) is not only adjustable (all defaults here), but even with default settings does NOT produce the awful "abrupt highlight clipping" as discussed in another thread, and mentioned by Adam Wilt (just see the extremely back-lit images at the bottom - the left one has been taken with Cine1 and shown in the "abrupt clipping" thread; see what I mean?).

So, the question is: to CINE, or not to CINE? - when one is after this punchy, contrasty and saturated look... Of course, the grabs on the right has been made too contrasty on purpose in order to prove my point; their blacks are too severely compressed for sure!

Piotr...

ALL the CINE gamma settings are designed for controlled lighting situations with fairly low contrast ranges. The STD gammas are more appropriate for scenes with wide dynamic range. CINE presets add a varying amount of black stretch at the expense of compressing the hi-lites. Not good for outdoor/sunny days with some shade in the scene. You seem to be somewhat resistant to realizing the effect of black stretch.

If I have been (am not any longer) "somewhat resistant to" anything, Bill, is the CINE settings not necessarily being more punchy than the standard ones (and this is because I got used to what CINE gamma / CINE color mean on the other prosumer Sony cameras, like the V1E I've been using so far). If you don't know what I am talking about, please see my threads in the V1 forum.

As to the effects of black stretch/compress, I am fully aware, and can take full and appropriate advantage of, Bill. If I used such low Black / Black Gamma values in the above examples, has only been to show how compressing blacks adds to the overall picture contrast - and I put a disclaimer about it!

...
As to the effects of black stretch/compress, I am fully aware, and can take full and appropriate advantage of. If I used such low Black / Black Gamma values has been to show how compressing blacks adds to the overall picture contrast.

Piotr:
I'd recommend shooting with Cinegamma (hisat matrix, Black and Blackgamma=0 and detail=off) and then adding Extra Punch in Post.

This is what I did with a simple avisynth script:


DirectShowSource("Path to your.MP4",fps=25,audio=false)
Tweak(sat=1.4,cont=1.2, bright=-10)
LimitedSharpen(ss_x=1.0,ss_y=1.0,Smode=3,strength=50)

Have you tried to add some punch in post?
It seems to work well, even with your "dark" pictures.

I did only use a S graduation curve on your picture ;)


regards Dennis

Dennis,

Everything can be done in post; I guess this is not what this particular thread is about, though:)

My point has been to convey to the other users who - like myself - might have upgraded from the cameras like the V1, that the CINE settings on the EX1 have quite different meaning: the contrasty and punchy look of the grabs I posted above with STD1 gamma, are very similar to what I've been getting with the Cine Gamma 2 on the V1E !

And - if you followed the long thread about the "abrupt highlights clipping" on this forum - to show that, just as Adam Wilt has noticed in his great EX1 review - the cine gammas' handling of it is somewhat tricky. This, you cannot make for in post!

Or if you think you can, please tell me how to repair the ugly blue patches in the sky as shown in my lower left grab above :)

So these blue batches also appear when using cinegamma (and not with STD Gamma with standard settings)?
Pretty strange that I've never noticed it, maybe it doesn't occur with mine, don't know.

I tried to reproduce this effect (ok, through a dirty window :D).
But it doesn't seem to appear...


Only CAs are more visible with STD gamma
First one is cine4, second is STD3,
not much difference, though

Yes - basing on my tests so far, the patches can appear with both CINE1 and STD3 (altough with STD3 it is possible to be avoided with its tweakable KNEE settings). Especially in the combination with the Hisat matrix.

On the other hand, the other gammas (I tested Cine4 thoroughly) have the knee low enough to prevent even getting near the offending highligts level, while the STD1 has it high enough (AND tweakable) to almost always be past it (i.e. safe on the right/high side).

Here goes a question that even after considerable experimentation time I am still unsure about the answer to:

- just what is the "Gamma Level" setting?

Unlike all the others, it's not easily trackable in the LCD by my naked eye at all. What is it supposed to do move the entire curve up or down? Or change its slope? Or move its main part, without touching the end points?

Gamma level shifts the position of middle gray. In other words, middle gray on the stock EX1, in Cine4, is about 55-60% IRE. A gamma level setting of -8 to -12 shifts middle gray to 50% IRE. The endpoints remain unchanged wjen gamma level is adjusted.

If one looks at a black step wedge on a WFM, middle gray is where the two steps intersect.

Look at Cine4:
http://www.dvinfo.net/gallery/showimage.php?i=851&c=2

Thanks Bill ;)

Gamma level shifts the position of middle gray. In other words, middle gray on the stock EX1, in Cine4, is about 55-60% IRE. A gamma level setting of -8 to -12 shifts middle gray to 50% IRE. The endpoints remain unchanged wjen gamma level is adjusted.

If one looks at a black step wedge on a WFM, middle gray is where the two steps intersect.

Look at Cine4:
http://www.dvinfo.net/gallery/showimage.php?i=851&c=2

When you say the foot intersects with the mid grey that seems to contradict what Sony says about the Cines being hyper-gammas which are constantly variable curves with no well defined turning point. In fact the Sony published curves show no foot?? But by now I think we know we have to work this out for ourselves. No one at SONY is being specific. Good we have you here Bill.

Did you see my "Political and Contraversial: NOT OT" post at the Vegas forum?


mike

Mike..

oh yeah
;o)


umm..did I say the foot intersects the mid grey? No, I said mid grey is where the left step wedge and right step wedge intersect.

Piotr...

ALL the CINE gamma settings are designed for controlled lighting situations with fairly low contrast ranges. The STD gammas are more appropriate for scenes with wide dynamic range. CINE presets add a varying amount of black stretch at the expense of compressing the hi-lites. Not good for outdoor/sunny days with some shade in the scene. You seem to be somewhat resistant to realizing the effect of black stretch.

I hope to do some major testing this weekend (if I get Cineform fixed) and hopefully I will develop an opinion of my own, but what you have said above is new to my ears and Bloom and Jenkins for two do otherwise.

I had just typed a lengthy response, but lost it as I was trying to post! Arrrggghh!

Anyway, to cut a long story short.
1. Cines 1 and 2 are best for film out (which was their primary purpose) and heavy grading.
2. Cines do not stretch the blacks. They compress the overload the chips are capable of capturing into the recordable signal range.
3. Contrary to what Bill wrote, the cine gammas are your friend in harsh sunny conditions.
4. Cines capture more contrast than STD gammas, but at the expense of tonal range.
5. STD gammas can be made to behave like Cines if you adjust knee point and knee slope, but the camera needs to be able to output a SAW signal so you can see the gamma curve on a waveform monitor.

Great points, Simon - fully backed up with what I experience. The idea of the "cine" look is not the "punchiness" and contrast at all - hence what we get with the Cine gammas is of a much less tonal range than the STD output. The only Cine gamma that stretches blacks (and blacks only) is Cine4.

This behaviour is inherited from the big CineAlta brothers, and is not quite consistent with what we saw in the prosumer cameras like the V1.

If one is after greater tonal range, especially in the highlights - he must use STD gammas (STD1 is by far the most dynamic and punchy) - but the "abrupt clipping" of colour may be more of a problem. Therefore I'd like to make some educated adjustment to the Knee settings of standard gammas, but for that I need some monitoring equipment - and here is my question: what do you mean by the SAW signal? Pardon my ignorance :)

The SAW signal is basically just a greyscale designed to represent the gamma curve that is output from the camera. When you adjust the knee, gamma etc you can see the gamma curve on the waveform monitor changing in realtime. In other words the SAW signal is designed so that you can actually see the gamma curve as a real time representation.

It isn't impossible to do without a SAW signal output, but it is more tricky.

I had just typed a lengthy response, but lost it as I was trying to post! Arrrggghh!

Anyway, to cut a long story short.
1. Cines 1 and 2 are best for film out (which was their primary purpose) and heavy grading.
2. Cines do not stretch the blacks. They compress the overload the chips are capable of capturing into the recordable signal range.
3. Contrary to what Bill wrote, the cine gammas are your friend in harsh sunny conditions.
4. Cines capture more contrast than STD gammas, but at the expense of tonal range.
5. STD gammas can be made to behave like Cines if you adjust knee point and knee slope, but the camera needs to be able to output a SAW signal so you can see the gamma curve on a waveform monitor.

Simon: Interesting post. May I suggest you go for a quick run up those Malvern hills, get some good clean air, and then sit down and write us the long version.

I hope in the long version you may more fully explain what you mean by heavy grading and tonal range as opposed to contrast. Why do you say only CINE 1 and 2 are good for film out?

I had just typed a lengthy response, but lost it as I was trying to post! Arrrggghh!When forming a long reply it's always best to write it up off-line (I use NotePad or WordPad) instead of the vBulletin interface, and then cut-and-paste into the Post Reply input field.

4. Cines capture more contrast than STD gammas, but at the expense of tonal range.

This one does not match what I am seeing, but it is likely only the wording that has me. STD settings seem to be capturing more contrast, the question is, does the scene allow for that to fit the dynamic range of the camera. Cines seem to be variations of contrast squashing.

This one does not match what I am seeing, but it is likely only the wording that has me. STD settings seem to be capturing more contrast, the question is, does the scene allow for that to fit the dynamic range of the camera. Cines seem to be variations of contrast squashing.

Exactly. That is one reason I asked for more explaination.

Ok, okay, it seems that we need a clarification of what is being talked about.

When I talk about contrast I am talking about the difference between light and dark. For example, in Standard gammas, if you have a bright sunny day, the shadows may look darker with crushed detail, while the highlights will be clipped. This will look more punchy, but detail will be lost. Hence some people may call this a high contrast look because darks are dark, and whites are very white. In actual fact it is a LOW contrast look because it is capturing limited detail in both the shadows and highlights. It can only capture decently either at the low to mid end, or the mid to high end, depending on your exposure.

The Cine gammas are high contrast because they can capture detail deep into the shadows AND detail into the highlights.

We need to remember that we are not discussing contrast in terms of how you might set your television set (ie a high contrast setting gives deep blacks and very white whites and therefore looks very punchy. I digress because contrast level on a TV set should really be called 'white level' :-) )

Cines seem to be variations of contrast squashing.

Technically in terms of how I am describing contrast you are correct. They are cramming maximum contrast range into the limited tonal range of the recorded signal.

Remember, that when you switch to, say Cine gamma 1 or 2, the darks may appear darker. But a lot more more highlight info is being crammed in there, so you can afford to bring the exposure up a bit to compensate.